Spies, Lies, and Weapons: What Went Wrong

How could we have been so far off in our estimates of Saddam Hussein's weapons programs? A leading Iraq expert and intelligence analyst in the Clinton Administration—whose book The Threatening Storm proved deeply influential in the run-up to the war—gives a detailed account of how and why we erred

By Kenneth M. Pollack
The Atlantic, January-February 2004

Let's start with one truth: last March, when the United States and its coalition partners invaded Iraq, the American public and much of the rest of the world believed that after Saddam Hussein's regime sank, a vast flotsam of weapons of mass destruction would bob to the surface. That, of course, has not been the case. In the words of David Kay, the principal adviser to the Iraq Survey Group (ISG), an organization created late last spring to search for prohibited weaponry, "I think all of us who entered Iraq expected the job of actually discovering deployed weapons to be easier than it has turned out to be." Many people are now asking very reasonable questions about why they were misled.



Democrats have typically accused the Bush Administration of exaggerating the threat posed by Iraq in order to justify an unnecessary war. Republicans have typically claimed that the fault lay with the CIA and the rest of the U.S. intelligence community, which they say overestimated the threat from Iraq—a claim that carries the unlikely implication that Bush's team might not have opted for war if it had understood that Saddam was not as dangerous as he seemed.

Both sides appear to be at least partly right. The intelligence community did overestimate the scope and progress of Iraq's WMD programs, although not to the extent that many people believe. The Administration stretched those estimates to make a case not only for going to war but for doing so at once, rather than taking the time to build regional and international support for military action.

This issue has some personal relevance for me. I began my career as a Persian Gulf military analyst at the CIA, where I saw an earlier generation of technical analysts mistakenly conclude that Saddam Hussein was much further away from having a nuclear weapon than the post-Gulf War inspections revealed. I later moved on to the National Security Council, where I served two tours, in 1995-1996 and 1999-2001. During the latter stint the intelligence community convinced me and the rest of the Clinton Administration that Saddam had reconstituted his WMD programs following the withdrawal of the UN inspectors, in 1998, and was only a matter of years away from having a nuclear weapon. In 2002 I wrote a book called Threatening Storm: The Case for Invading Iraq, in which I argued that because all our other options had failed, the United States would ultimately have to go to war to remove Saddam before he acquired a functioning nuclear weapon. Thus it was with more than a little interest that I pondered the question of why we didn't find in Iraq what we were so certain we would.

What We Thought We Knew

The U.S. intelligence community's belief that Saddam was aggressively pursuing weapons of mass destruction pre-dated Bush's inauguration, and therefore cannot be attributed to political pressure. It was first advanced at the end of the 1990s, at a time when President Bill Clinton was trying to facilitate a peace agreement between Israel and the Palestinians and was hardly seeking assessments that the threat from Iraq was growing.

In congressional testimony in March of 2002 Robert Einhorn, Clinton's assistant secretary of state for nonproliferation, summed up the intelligence community's conclusions about Iraq at the end of the Clinton Administration:

"How close is the peril of Iraqi WMD? Today, or at most within a few months, Iraq could launch missile attacks with chemical or biological weapons against its neighbors (albeit attacks that would be ragged, inaccurate, and limited in size). Within four or five years it could have the capability to threaten most of the Middle East and parts of Europe with missiles armed with nuclear weapons containing fissile material produced indigenously—and to threaten U.S. territory with such weapons delivered by nonconventional means, such as commercial shipping containers. If it managed to get its hands on sufficient quantities of already produced fissile material, these threats could arrive much sooner."

In October of 2002 the National Intelligence Council, the highest analytical body in the U.S. intelligence community, issued a classified National Intelligence Estimate on Iraq's WMD, representing the consensus of the intelligence community. Although after the war some complained that the NIE had been a rush job, and that the NIC should have been more careful in its choice of language, in fact the report accurately reflected what intelligence analysts had been telling Clinton Administration officials like me for years in verbal briefings.

A declassified version of the 2002 NIE was released to the public in July of last year. Its principal conclusions:

* "Iraq has continued its weapons of mass destruction (WMD) programs in defiance of UN resolutions and restrictions. Baghdad has chemical and biological weapons as well as missiles with ranges in excess of UN restrictions; if left unchecked, it probably will have a nuclear weapon during this decade." (The classified version of the NIE gave an estimate of five to seven years.)

* "Since inspections ended in 1998, Iraq has maintained its chemical weapons effort, energized its missile program, and invested more heavily in biological weapons; most analysts assess [that] Iraq is reconstituting its nuclear weapons program."

* "If Baghdad acquires sufficient weapons-grade fissile material from abroad, it could make a nuclear weapon within a year ... Without such material from abroad, Iraq probably would not be able to make a weapon until the last half of the decade."

* "Baghdad has begun renewed production of chemical warfare agents, probably including mustard, sarin, cyclosarin, and VX ... Saddam probably has stocked a few hundred metric tons of CW agents."

* "All key aspects—R&D, production, and weaponization—of Iraq's offensive BW [biological warfare] program are active and most elements are larger and more advanced than they were before the Gulf war ... Baghdad has established a large-scale, redundant, and concealed BW agent production capability, which includes mobile facilities; these facilities can evade detection, are highly survivable, and can exceed the production rates Iraq had prior to the Gulf war."

U.S. government analysts were not alone in these views. In the late spring of 2002 I participated in a Washington meeting about Iraqi WMD. Those present included nearly twenty former inspectors from the United Nations Special Commission (UNSCOM), the force established in 1991 to oversee the elimination of WMD in Iraq. One of the senior people put a question to the group: Did anyone in the room doubt that Iraq was currently operating a secret centrifuge plant? No one did. Three people added that they believed Iraq was also operating a secret calutron plant (a facility for separating uranium isotopes).

Other nations' intelligence services were similarly aligned with U.S. views. Somewhat remarkably, given how adamantly Germany would oppose the war, the German Federal Intelligence Service held the bleakest view of all, arguing that Iraq might be able to build a nuclear weapon within three years. Israel, Russia, Britain, China, and even France held positions similar to that of the United States; France's President Jacques Chirac told Time magazine last February, "There is a problem—the probable possession of weapons of mass destruction by an uncontrollable country, Iraq. The international community is right ... in having decided Iraq should be disarmed." In sum, no one doubted that Iraq had weapons of mass destruction.

What We Think We Know Now

But it appears that Iraq may not have had any actual weapons of mass destruction. A number of caveats are in order. We do not yet have a complete picture of Iraq's WMD programs. Initial U.S. efforts to seek out WMD caches were badly lacking: an American artillery unit that had too few people for the task and virtually no plan of action had been hastily assigned the mission. Not surprisingly, its efforts garnered little useful information. According to Judith Miller, a New York Times reporter who was embedded with the unit, by mid-June—nearly two months after the end of major combat operations—the United States had interviewed only thirteen out of hundreds of Iraqi scientists. Documents relating to the programs are known to have been destroyed. Much of Iraq is yet to be explored; as David Kay, of the Iraq Survey Group, which took over the search for WMD in June, told Congress, only ten of Iraq's 130 major ammunition dumps had been thoroughly checked as of early October (the time of his testimony). Now that Saddam Hussein is in custody, it is possible that new information may be forthcoming, or that closemouthed Iraqis will offer fresh details.

Nevertheless, the preliminary findings of the ISG will probably not change dramatically, at least not in their broad contours. Kay summarized those findings in his October testimony.

* Iraq had preserved some of its technological nuclear capability from before the Gulf War. However, no evidence suggested that Saddam had undertaken any significant steps after 1998 toward reconstituting the program to build nuclear weapons or to produce fissile material.

* Little evidence surfaced that Iraq had continued to produce chemical weapons; only a minimal amount of clandestine research had been done on them. For instance, the production line at the Fallujah II facility (the plant that intelligence officers believed was Iraq's principal site for making chlorine, an ingredient in some chemical-warfare agents) turned out to be in derelict condition and had not operated since the Gulf War. Nevertheless, Iraqi officials seemed to believe that they could convert existing civilian pharmaceutical plants to chemical-weapons production, and that Saddam was interested in their ability to do so.

* Iraq made determined efforts to retain some capabilities for biological warfare. It maintained an undeclared network of laboratories and other facilities within the apparatus of its security services, and as Kay put it, "this clandestine capability was suitable for preserving BW expertise, BW-capable facilities, and continuing R&D—all key elements for maintaining a capability for resuming BW production." To disguise its biological-warfare programs Baghdad had scientists working on overt projects that were closely related to proscribed activities.

* Iraq seemed to have been most aggressive in pursuing proscribed missiles. In Kay's words, "detainees and cooperative sources indicate that beginning in 2000 Saddam ordered the development of ballistic missiles with ranges of at least [240 miles] and up to [620 miles] and that measures to conceal these projects from [UN inspectors] were initiated in late 2002, ahead of the arrival of inspectors." The Iraqis were also working on clustering liquid-fueled rocket engines in order to produce a longer-range missile, and were trying to convert certain surface-to-air missiles into surface-to-surface missiles with a range of 150 miles. Most troubling of all, the ISG uncovered evidence that from 1999 to 2002 Iraq had negotiated with North Korea to buy technology for No Dong missiles, which have a range of 800 miles.

Overall, these findings suggest that Iraq did retain prohibited WMD programs, but that those programs were not so extensive, advanced, or threatening as the National Intelligence Estimate maintained.

More-cautious analysts had argued that the NIE's assessment that Iraq had large stockpiles of chemical and biological weapons was unlikely, because such munitions deteriorate rapidly and can be quickly produced in bulk if production lines and precursor agents are available (making stockpiles unnecessary as well as inefficient). These analysts instead believed that Iraq had a "just-in-time" production capability—that it could churn out these weapons as needed, using hidden or dual-use facilities. But not even this more conservative scenario was borne out by the ISG's investigations. Sources told the group that Saddam and his son Uday had each, on separate occasions in 2001 and 2002, asked officials associated with Iraq's chemical-warfare program how long it would take to produce chemical agents and weapons. One official reportedly told Saddam that it would take six months to produce mustard gas (among the easiest such agents to manufacture); another told Uday that it would take two months to produce mustard gas and two years to produce sarin (a simple nerve agent). The questions do not suggest the presence of large stockpiles. The answers do not support a just-in-time capability.

The ISG's findings to date are most damning in the nuclear arena—as it happens, the segment of Iraq's WMD program in which the initial findings are most likely to be correct, because nuclear-weapons production is extremely difficult to conceal. The perceived nuclear threat was always the most disturbing one. The U.S. intelligence community's belief toward the end of the Clinton Administration that Iraq had reconstituted its nuclear program and was close to acquiring nuclear weapons led me and other Administration officials to support the idea of a full-scale invasion of Iraq, albeit not right away. The NIE's judgment to the same effect was the real linchpin of the Bush Administration's case for an invasion.

What we have found in Iraq since the invasion belies that judgment. Saddam did retain basic elements for a nuclear-weapons program and the desire to acquire such weapons at some point, but the program itself was dormant. Saddam had not ordered its resumption (although some reports suggest that he considered doing so in 2002). In all probability Iraq was considerably further from having a nuclear weapon than the five to seven years estimated in the classified version of the NIE.

The View From Baghdad

F iguring out why we overestimated Iraq's WMD capabilities involves figuring out what the Iraqis, especially Saddam Hussein, were thinking and doing throughout the 1990s. The story starts right after the Gulf War. An Iraqi document that fell into the inspectors' hands revealed that in April of 1991 a high-level Iraqi committee had ordered many of the country's WMD activities to be hidden from UN inspectors, even though compliance with the inspections was a condition for the lifting of economic sanctions imposed after the invasion of Kuwait. The document was a report from a nuclear-weapons plant describing how it carried out this order. According to UNSCOM's final report, "The facility was instructed to remove evidence of the true activities at the facility, evacuate documents to hide sites, make physical alterations to the site to hide its true purpose, develop cover stories, and conduct mock inspections to prepare for UN inspectors."

A great deal of other information substantiates the idea that Saddam at first decided to try to keep a considerable portion of his WMD programs intact and hidden. His efforts probably included retaining some munitions, but mainly concerned production and research elements. In other words, Saddam did initially try to maintain a "just-in-time" capability. However, it became increasingly clear how difficult this would be. In the summer of 1991 inspectors tracked down and destroyed Saddam's calutrons. Their discoveries may have convinced him that he would have to put his WMD programs on hold until after the sanctions were lifted—something he reportedly thought would happen within a matter of months.

But the inspectors proved more tenacious and the international community more steadfast than the Iraqis had expected. Accordingly, from June of 1991 to May of 1992 Iraq unilaterally destroyed parts of its WMD programs (as we know from subsequent Iraqi admissions). This action appears to have served two purposes: It got rid of unnecessary munitions and secondary equipment that the inspectors might have found, which would have constituted proof of Iraqi noncompliance. And it helped Baghdad conceal more-important elements of the programs, because the regime could point to the unilateral destructions as evidence of cooperation and could claim that even more material had been destroyed. (Since the fall of Baghdad scientists have told the ISG that key equipment was in fact diverted from these destructions and hidden.)

In 1995 matters changed. That August, Hussein Kamel, Saddam's son-in-law and the head of Iraq's WMD programs, defected to Jordan, prompting a panicked Baghdad to hurriedly turn over hundreds of thousands of pages of new documentation to the United Nations. According to the former chief UN weapons inspector Rolf Ekeus, Kamel's statements and the Iraqi documents squared with what UNSCOM had been finding: although all actual weapons had been eliminated, either by the UN or in the earlier destructions, Iraq had preserved production and R&D programs. Although the Iraqis tried to withhold any highly incriminating documents from the UN (and, ridiculously, claimed that Kamel had been running the programs on his own, without anyone else's knowledge), in their rush they overlooked several containing crucial information about previously concealed aspects of the nuclear and biological programs.

Other secrets were laid bare that same year. A U.S.-UN sting operation caught the Iraqis trying to smuggle 115 missile gyroscopes through Jordan. (UN inspectors later found other gyroscopes hidden at the bottom of the Tigris River.) Iraq was forced to admit to the existence of a facility to build Scud-missile engines, and to destroy a hidden plant for manufacturing modified Scud missiles. It was also forced to admit to having made much greater progress on its nuclear program before the Gulf War than it had previously acknowledged. Most important, it was forced to admit that a very large biological-weapons plant at al-Hakim, whose existence had been concealed from UN inspectors, had produced 500,000 liters of biological agents in 1989 and 1990, and that it was still functional in 1995. Three years after this confession Lieutenant General Amer al-Saadi, Saddam's principal liaison with the UN, told inspectors that Iraq would offer no excuse or defense for having denied the existence of its biological-weapons program. He stated matter-of-factly that Iraq had made a political decision to conceal it.

Either late in 1995 or at some point in 1996 Saddam probably recognized that trying to retain his just-in-time capability had become counterproductive. The inspectors kept finding pieces of the programs, and each discovery pushed the lifting of the sanctions further into the future. It's important to keep in mind several other events of this period. Saddam's internal position was very shaky. He had faced disturbances in several of his most loyal Sunni tribes. In addition to Kamel, a number of high-ranking officials had defected to the West, including Saddam's chief of military intelligence, Wafic Samarai. Coup plots abounded. In 1995 the Kurds smashed two Iraqi infantry brigades at Irbil, humiliating the Iraqi army. In 1996 Iraqi intelligence uncovered a CIA-backed coup attempt whose participants had penetrated some of Saddam's most sensitive intelligence services. Iraq's economy was suffocating under the sanctions, and inflation was rampant. Given this precarious situation, Saddam probably decided to scale back his WMD programs (with the likely exception of work on proscribed missiles, which could be concealed by Iraq's permitted missile program) by destroying additional equipment, keeping the bare minimum needed to rebuild them at some point, in order to reduce the risk of further discoveries. This would have meant giving up the idea of just-in-time production capabilities and limiting his efforts to hiding documents and only key pieces of equipment. In short, Saddam switched from trying to hang on to the maximum production and research assets of his WMD programs to trying to keep only the minimum necessary to reconstitute the programs at some point after the sanctions had been lifted.

What Was Saddam Thinking?

Having decided to give up so much of his WMD capability, why didn't Saddam change his behavior toward the UN inspectors and demonstrate a spirit of candor and cooperation? Even after 1996 the Iraqis took a confrontational posture toward UNSCOM, fighting to prevent inspectors from going where they wanted to go and seeing what they wanted to see. The governments of the world inferred from this defiance that Saddam was still not complying with the UN resolutions, and the sanctions therefore stayed in place.

The first and most obvious answer is that Saddam still had some things to hide, and was fearful of their discovery. Although he did unquestionably have some things to hide, this answer is not entirely satisfying. Iraq was able to conceal the minimized remnants of its WMD programs so well that UNSCOM found little incriminating evidence in 1997 and 1998. This early success should have given Saddam the confidence to begin to cooperate more fully with the UN resolutions. But throughout the period leading up to the war Saddam remained as obstinate as ever.

An alternative explanation, offered by Iraq's former UN ambassador, Tariq Aziz, and other officials captured after last year's war, goes like this: Saddam was pretending to have WMD in order to enhance his prestige among the other Arab nations. This explanation doesn't ring completely true either. It is certainly the case that Saddam garnered a great deal of admiration from Arabs of many countries by appearing to have such weapons, and that he aspired to dominate the Arab world. But this theory assumes that he was willing to incur severe penalties for the UN's belief that he still had WMD without reaping any tangible benefits from actually having them. If prestige had been more important to him than the lifting of the sanctions, it would have been more logical and more in keeping with his character to simply retain all his WMD capabilities.

Saddam's behavior may have been driven by completely different considerations. Saddam has always evinced much greater concern for his internal position than for his external status. He has made any number of highly foolish foreign-policy decisions—for example, invading Kuwait and then deciding to stick around and fight the U.S.-led coalition—in response to domestic problems that he feared threatened his grip on power. The same forces may have been at work here; after all, ever since the Iran-Iraq war WMD had been an important element of Saddam's strength within Iraq. He used them against the Kurds in the late 1980s, and during the revolts that broke out after the Gulf War, he sent signals that he might use them against both the Kurds and the Shiites. He may have feared that if his internal adversaries realized that he no longer had the capability to use these weapons, they would try to move against him. In a similar vein, Saddam's standing among the Sunni elites who constituted his power base was linked to a great extent to his having made Iraq a regional power—which the elites saw as a product of Iraq's unconventional arsenal. Thus openly giving up his WMD could also have jeopardized his position with crucial supporters.

Furthermore, Saddam may have felt trapped by his initial reckoning that he could fool the UN inspectors and that the sanctions would be short-lived. Because of this mistaken calculation he had subjected Iraq to terrible hardships. Suddenly cooperating with the inspectors would have meant admitting to both his opponents and his supporters that his course of action had been a mistake and that, having now given up most of his WMD programs, he had devastated Iraqi society for no reason.

This suggests that in 1995-1996 Saddam took one of his famous gambles—gambles that almost never worked out for him. He chose not to "come clean" and cooperate with the UN for fear that this would make him look weak to both his domestic enemies and his domestic allies, either of whom might then have moved against him. But he would try to greatly diminish the chances that UNSCOM would find more evidence of his continuing noncompliance by reducing his WMD programs to the bare minimum, in hopes that the absence of evidence would lead to the lifting of sanctions—something he desperately sought in 1996.

In other respects Saddam's fortunes began to rise in 1996. Although the CIA-backed coup attempt may have signified internal weakness, the fact that Saddam snuffed it out, as he had many previous attempts, signified strength. Also, to avenge the Iraqi army's 1995 defeat at Irbil, Saddam manipulated infighting among the Kurds so as to allow his Republican Guards to drive into the city, smash the Kurd defenders, and arrest several hundred CIA-backed rebels. As the historian Amatzia Baram has persuasively argued in his book Building Toward Crisis (1998), these successes made Saddam feel secure enough to swallow his pride and accept UN Resolution 986, the oil-for-food program, which he had previously rejected as an infringement on Iraqi sovereignty. Oil-for-food turned out to be an enormous boon for the Iraqi economy, and commodity prices fell quickly, stabilizing the dinar.

The oil-for-food program itself gave Saddam clout to apply toward the lifting of the sanctions. Under Resolution 986 Iraq could choose to whom it would sell its oil and from whom it would buy its food and medicine. Baghdad could therefore reward cooperative states with contracts. Not surprisingly, France and Russia regularly topped the list of Iraq's oil-for-food partners. In addition, Iraq could set the prices—and since Saddam did not really care whether he was importing enough food and medicine for his people's needs, he could sell oil on the cheap and buy food and medicine at inflated prices as additional payoff to friendly governments. He made it clear that he wanted his trading partners to ignore Iraqi smuggling and try to get the sanctions lifted.

By 1997 the international environment had changed markedly, in ways that probably convinced Saddam that he didn't need to cooperate with the inspectors. The same international outcry—against the suffering inflicted by the Iraq sanctions—that prompted the United States to craft the oil-for-food deal was creating momentum for lifting the sanctions completely. At that point it was reasonable for Saddam to believe that in the not too distant future the sanctions either would be lifted or would be so undermined as to be effectively meaningless, and that he would never have to reveal the remaining elements of his WMD programs. Only in 2002, when the Bush Administration suddenly focused its attention on Iraq, would Saddam have had any reason to change this view. And then, according to a variety of Iraqi sources, he simply refused to believe that the Americans were serious and would actually invade.

Another explanation should be posited. This is the notion that Saddam did not order the program scaled down, but Iraqi scientists ensured that it did not progress and deceived Saddam into believing that it was much further along than it in fact was. Numerous Iraqi scientists have claimed that although Saddam ordered them to produce particular things for the WMD programs, they dragged their feet or found other ways to avoid delivering them. There is most likely a germ of truth to these stories: prevarication on the part of some Iraqi scientists may have helped to account for the modest state of Iraq's WMD programs in 2003. But they probably form only a part of the explanation. Many of the accounts of scientists' quietly thwarting Saddam are undoubtedly self-serving, concocted in the aftermath of his defeat. As we have heard time and again from Iraqi defectors, those who did not meet Saddam's demands risked torture and murder for themselves and their families. We have consistently found that in Saddam's Iraq very few people took that risk.

One last element may also have been at work all along: the possibility that Saddam genuinely feared that the inspections were a cover for a CIA campaign to overthrow or assassinate him. The Iraqis repeatedly cited this fear in denying UNSCOM access to certain "sensitive" sites—particularly palaces—that were associated with Saddam personally. The rest of the world assumed that it was merely an excuse to keep inspectors out of places that contained evidence of WMD programs. However, the Iraqis may have been telling the truth on this point (and the initial debriefing of Saddam lends some credence to this scenario). After all, as various sources have now disclosed, the United States did run a covert-action campaign against Saddam, starting in 1991, and U.S. intelligence did use UNSCOM operations (without UNSCOM's knowledge) to gather intelligence for that campaign.

The Perils of Prediction

Everyone outside Iraq missed the 1995-1996 shift in Saddam's strategy—that is, to scale back his WMD programs to minimize the odds of further discoveries—and assumed that Iraq's earlier behavior was continuing more or less in a straight line. This misperception took on considerable weight in the following years.

Context is crucial to understanding any intelligence assessment. No matter how objective the analyst may be, he or she begins with a set of basic assumptions that create a broad perspective on an issue; this helps the analyst to sort through evidence.

The context for the 2002 NIE assessment of Iraq's WMD programs began to take shape before the Gulf War. Prior to 1991 the intelligence communities in the United States and elsewhere believed that Iraq was at least five, and probably closer to ten, years away from acquiring a nuclear weapon. Of course, after the war we learned that in 1991 Iraq had been only six to twenty-four months away from having a workable nuclear weapon. This revelation stunned the analysts responsible for following the Iraqi nuclear program. The lessons they took from it were that Iraq was determined to acquire nuclear weapons and would go to any lengths to do so; that in pursuit of this goal Iraq was willing to use technology that Westerners considered crude and obsolete; that the Iraqis were superb at concealment and deception; and that inspections were inherently flawed—after all, there had been inspectors in Iraq prior to 1990, and they had been completely fooled.

These lessons were strongly reinforced by the revelation of Iraq's attempts in the first four years after the war to preserve significant parts of its WMD programs. By about 1994 UNSCOM believed, incorrectly, that it had largely disarmed Iraq; its members were privately discussing switching its operations from active inspections to passive monitoring. Many intelligence analysts in the United States, Britain, and Israel disagreed with UNSCOM's assessment, but they were hard-pressed to substantiate their suspicions—until Hussein Kamel's defection, in 1995, and subsequent Iraqi admissions regarding the extent of deception. These developments came as a profound shock to the UN inspectors, who resolved that Iraq could never again be trusted. Thus, just when Iraq was in all likelihood giving up efforts to maintain its just-in-time production capability, the rest of the world became hardened in its conviction that Saddam would never abandon or even reduce his efforts to acquire WMD.

Another important contribution to the context is the continuation of Saddam's hostility toward the inspectors. If anything, the Iraqis became even less accommodating over time. By 1998 they were physically harassing the inspectors—on one occasion firing two rocket-propelled grenades into an UNSCOM building in Baghdad, on another grabbing the controls of an UNSCOM helicopter in flight and nearly causing it to crash. Western intelligence agencies understandably took these actions to mean that nothing in Saddam's weaponry plans had changed.

In December of 1998 the inspectors withdrew from the country. Their decision to do so came after Iraq announced, in August of that year, that it would no longer cooperate with them at all, and after repeated crises demonstrated that Baghdad's announcement was not just bluster.

The end of the UN inspections appears in retrospect to have been a much greater problem than anyone recognized at the time. The inspectors had been the best source of information on Iraq and its WMD programs. UNSCOM had a large and highly capable cadre of weapons specialists who focused exclusively on Iraq. Many Western intelligence agencies, faced with other issues that demanded their resources, increasingly relied on UNSCOM's data and assessments and did little to bolster their own (meager) capabilities in Iraq. And UNSCOM had something that American intelligence did not—physical access to Iraq. Without an embassy there it was very hard for U.S. case officers to penetrate the country.

The end of the inspections eliminated the single best means of vetting what information intelligence agencies could gather independently about Iraq. These agencies usually shared (in some form) new information or analyses about the WMD programs with UNSCOM. If a defector claimed that biological-weapons material was stored at a given site, inspectors would look for it. If satellite imagery indicated unusual activity at a particular location, inspectors would try to confirm it. Although Iraq's counterintelligence efforts were formidable (UNSCOM estimated that only six of its roughly 250 inspections actually caught the Iraqis by surprise), UNSCOM was usually able to gauge, if only broadly, whether a source or a deduction was correct.

When the inspectors suddenly left, the various intelligence agencies were caught psychologically and organizationally off balance. Desperate for information on Iraq, they began to trust sources that they would previously have had UNSCOM vet. If a defector came out of Iraq after 1998, the CIA had to gauge his credibility by comparing his account with those of other defectors—who might be unreliable or just unproven—or by checking it against whatever they could glean from satellites and other indirect sources. With so little to go on, intelligence agencies believed many reports that now seem deeply suspect.

In the absence of hard evidence, the intelligence analysts tended to fall back on the underlying assumptions they had begun with. Those assumptions included the belief that Saddam was determined to preserve his extant WMD capabilities and acquire new ones. And now there were no weapons inspectors to hinder him. The inspectors had also been a moderating influence on Western intelligence agencies; the information they provided, and the mere fact of their presence in Iraq, helped those agencies stick to reasonable suppositions and keep unsubstantiated fears at bay. After 1998 many analysts increasingly entertained worst-case scenarios—scenarios that gradually became mainstream estimates.

Another element that contributed to faulty assessments before the 2003 invasion was Iraqi rhetoric. Imagine that you were a CIA analyst in June of 2000 and heard Saddam make the following statement: "If the world tells us to abandon all our weapons and keep only swords, we will do that. We will destroy all the weapons, if they destroy their weapons. But if they keep a rifle and then tell me that I have the right to possess only a sword, then we would say no. As long as the rifle has become a means to defend our country against anybody who may have designs against it, then we will try our best to acquire the rifle." It would be very difficult not to interpret Saddam's remarks as an announcement that he intended to reconstitute his WMD programs.

The final element in the context for our pre-invasion analysis involved discrepancies between how much WMD material went into Iraq and how much Iraq could prove it had destroyed. Before the Gulf War (and to a certain extent afterward) Baghdad imported enormous quantities of equipment and raw materials for WMD. The UN inspectors, with remarkable diligence, obtained virtually all the import figures, either from the Iraqis or from their suppliers. They then asked the Iraqis to either produce the materials or account for their destruction. In many cases the Iraqis could not. The difference between what they had imported and what they could account for was seen as important evidence of an ambitious clandestine WMD program. These are the numbers—of bombs, of liters of precursor chemicals, and so on—that the world regularly heard Bush Administration officials intone during the run-up to the 2003 war.

In hindsight there are legitimate reasons to question these numbers. According to David Kay, a number of Iraqi sources have told the ISG that some of the material that was unaccounted for was diverted from the unilateral destructions that took place from 1991 to 1996. However, it is not clear whether or not any of that material was destroyed later. And it is likely that some of the discrepancies between UNSCOM and Iraqi figures are no more than the result of sloppiness. Saddam's Iraq was not exactly an efficient state, and many of his chief lieutenants were semi-literate thugs with no understanding of esoteric technical matters and little regard for how things should be done—their only concern was that Saddam's demands be met.

The Politics of Persuasion

The intelligence community's overestimation of Iraq's WMD capability is only part of the story of why we went to war last year. The other part involves how the Bush Administration handled the intelligence. Throughout the spring and fall of 2002 and well into 2003 I received numerous complaints from friends and colleagues in the intelligence community, and from people in the policy community, about precisely that. According to them, many Administration officials reacted strongly, negatively, and aggressively when presented with information or analysis that contradicted what they already believed about Iraq. Many of these officials believed that Saddam Hussein was the source of virtually all the problems in the Middle East and was an imminent danger to the United States because of his perceived possession of weapons of mass destruction and support of terrorism. Many also believed that CIA analysts tended to be left-leaning cultural relativists who consistently downplayed threats to the United States. They believed that the Agency, not the Administration, was biased, and that they were acting simply to correct that bias.

Intelligence officers who presented analyses that were at odds with the pre-existing views of senior Administration officials were subjected to barrages of questions and requests for additional information. They were asked to justify their work sentence by sentence: "Why did you rely on this source and not this other piece of information?" "How does this conclusion square with this other point?" "Please explain the history of Iraq's association with the organization you mention in this sentence." Reportedly, the worst fights were those over sources. The Administration gave greatest credence to accounts that presented the most lurid picture of Iraqi activities. In many cases intelligence analysts were distrustful of those sources, or knew unequivocally that they were wrong. But when they said so, they were not heeded; instead they were beset with further questions about their own sources.

On many occasions Administration officials' requests for additional information struck the analysts as being made merely to distract them from their primary mission. Some officials asked for extensive historical analyses—a hugely time-consuming undertaking, for which most intelligence analysts are not trained. Requests were constantly made for detailed analyses of newspaper articles that conformed to the views of Administration officials—pieces by conservative newspaper columnists such as Jim Hoagland, William Safire, and George F. Will. These columnists may be highly intelligent men, but they have no claim to superior insight into the workings of Iraq, or to any independent intelligence-collection capabilities.

Of course, no policymaker should accept intelligence estimates unquestioningly. While I was at the NSC, I regularly challenged analysts as to why they believed what they did. I asked for additional material and required them to do significant additional work. Any official who does less is derelict in his or her duty. However, at a certain point curiosity and diligence become a form of pressure. If your employer asks you every so often about your health and seems to take an appropriate interest in the answer, you probably feel that he or she is kind and considerate. If your employer asks you about your health every ten minutes, in highly detailed, probing questions, you may have a more nervous reaction.

As Seymour Hersh, among others, has reported, Bush Administration officials also took some actions that arguably crossed the line between rigorous oversight of the intelligence community and an attempt to manipulate intelligence. They set up their own shop in the Pentagon, called the Office of Special Plans, in order to sift through the information on Iraq themselves. To a great extent OSP personnel "cherry-picked" the intelligence they passed on, selecting reports that supported the Administration's pre-existing position and ignoring all the rest.

Most problematic of all, the OSP often chose to believe reports that trained intelligence officers considered unreliable or downright false. In particular it gave great credence to reports from the Iraqi National Congress, whose leader was the Administration-backed Ahmed Chalabi. It is true that the intelligence community believed some of the material that came from the INC—but not most of it. (In retrospect, of course, it seems that even the intelligence professionals gave INC reporting more credence than it deserved.) One of the reasons the OSP generally believed Chalabi and the INC was that they were telling it what it wanted to hear—giving the OSP, in a kind of vicious circle, further incentive to trust these sources over differing, and ultimately more reliable, ones. Thus intelligence analysts spent huge amounts of time fighting bad information and trying to persuade Administration officials not to make policy decisions based on it. From my own experience I know that it is hard enough to figure out what the reliable evidence indicates—and vast battles are fought over that. To have to also fight over what is clearly bad information is a Sisyphean task.

The Bush officials who created the OSP gave its reports directly to those in the highest levels of government, often passing raw, unverified intelligence straight to the Cabinet level as gospel. Senior Administration officials made public statements based on these reports—reports that the larger intelligence community knew to be erroneous (for instance, that there was hard and fast evidence linking Iraq to al-Qaeda). Another problem arising from the machinations of the OSP is that whenever the principals of the National Security Council met with the President and his staff, two completely different versions of reality were on the table. The CIA, the State Department, and the uniformed military services would present one version, consistent with the perspective of intelligence and foreign-policy professionals, and the Office of the Secretary of Defense and the Office of the Vice President would present another, based on the perspective of the OSP. These views were too far apart to allow for compromise. As a result, the Administration found it difficult, if not impossible, to make certain important decisions. And it made some that were fatally flawed, including many relating to postwar planning, when the OSP's view—that Saddam's regime simultaneously was very threatening and could easily be replaced by a new government—prevailed.

For the most part, the problems discussed so far have more to do with the methods of Administration officials than with their motives, which were often misguided and dangerous, but were essentially well-intentioned. The one action for which I cannot hold Administration officials blameless is their distortion of intelligence estimates when making the public case for going to war.

As best I can tell, these officials were guilty not of lying but of creative omission. They discussed only those elements of intelligence estimates that served their cause. This was particularly apparent in regard to the time frame for Iraq's acquisition of a nuclear weapon—the issue that most alarmed the American public and the rest of the world. Remember that the NIE said that Iraq was likely to have a nuclear weapon in five to seven years if it had to produce the fissile material indigenously, and that it might have one in less than a year if it could obtain the material from a foreign source. The intelligence community considered it highly unlikely that Iraq would be able to obtain weapons-grade material from a foreign source; it had been trying to do so for twenty-five years with no luck. However, time after time senior Administration officials discussed only the worst-case, and least likely, scenario, and failed to mention the intelligence community's most likely scenario. Some examples:

* In a radio address on September 14, 2002, President Bush warned, "Today Saddam Hussein has the scientists and infrastructure for a nuclear-weapons program, and has illicitly sought to purchase the equipment needed to enrich uranium for a nuclear weapon. Should his regime acquire fissile material, it would be able to build a nuclear weapon within a year."

* On October 7, 2002, the President told a group in Cincinnati, "If the Iraqi regime is able to produce, buy, or steal an amount of highly enriched uranium a little larger than a single softball, it could have a nuclear weapon in less than a year."

* On November 1, 2002, Undersecretary of State John Bolton told the Second Global Conference on Nuclear, Bio/Chem Terrorism, "We estimate that once Iraq acquires fissile material—whether from a foreign source or by securing the materials to build an indigenous fissile-material capability—it could fabricate a nuclear weapon within one year."

* Vice President Cheney said on NBC's Meet the Press on September 14, 2003, "The judgment in the NIE was that if Saddam could acquire fissile material, weapons-grade material, that he would have a nuclear weapon within a few months to a year. That was the judgment of the intelligence community of the United States, and they had a high degree of confidence in it."

None of these statements in itself was untrue. However, each told only a part of the story—the most sensational part. These statements all implied that the U.S. intelligence community believed that Saddam would have a nuclear weapon within a year unless the United States acted at once.

Some defenders of the Administration have reportedly countered that all it did was make the best possible case for war, playing a role similar to that of a defense attorney who is charged with presenting the best possible case for a client (even if the client is guilty). That is a false analogy. A defense attorney is responsible for presenting only one side of a dispute. The President is responsible for serving the entire nation. Only the Administration has access to all the information available to various agencies of the U.S. government—and withholding or downplaying some of that information for its own purposes is a betrayal of that responsibility.

What Is to Be Done?

What we have learned about Iraq's WMD programs since the fall of Baghdad leads me to conclude that the case for war with Iraq was considerably weaker than I believed beforehand. Because of the consensus among American and foreign intelligence agencies, outside experts, and former UN weapons inspectors, I had been convinced that Iraq was only years away from having a nuclear weapon—probably only four or five years, as Robert Einhorn had testified. That estimate was clearly off, possibly by quite a bit. My reluctant conviction that war was our only option (although not at the time or in the manner in which the Bush Administration pursued it) was not entirely based on the nuclear threat, but that threat was the most important factor in it.

The war was not all bad. I do not believe that it was a strategic mistake, although the appalling handling of postwar planning was. There is no question that Saddam Hussein was a force for real instability in the Persian Gulf, and that his removal from power was a tremendous improvement. There is also no question that he was pure evil, and that he headed one of the most despicable regimes of the past fifty years. I am grateful that the United States no longer has to contend with the malign influence of Saddam's Iraq in this economically irreplaceable and increasingly fragile part of the world; nor can I begrudge the Iraqi people one day of their freedom. What's more, we should not forget that containment was failing. The shameful performance of the United Nations Security Council members (particularly France and Germany) in 2002-2003 was final proof that containment would not have lasted much longer; Saddam would eventually have reconstituted his WMD programs, although further in the future than we had thought. That said, the case for war—and for war sooner rather than later—was certainly less compelling than it appeared at the time. At the very least we should recognize that the Administration's rush to war was reckless even on the basis of what we thought we knew in March of 2003. It appears even more reckless in light of what we know today.

The problems that led to our mistaken beliefs about the threat posed by Iraq's weapons of mass destruction must be addressed immediately. Unfortunately, to some extent the problems are contradictory, and therefore the solutions may work against one another. For example, a remedy used in the past to address influence from the executive branch on the intelligence process has been to increase oversight of intelligence operations and analysis by Congress. However, in this instance increasing congressional oversight could have exacerbated another problem: the failure of the intelligence community to sufficiently challenge its own assumptions about Saddam's strategy. The more that intelligence agencies must report to both Congress and the White House, the more they fear becoming a political football, and the more they will tone down their estimates, stick to mainstream judgments, and avoid taking controversial positions. Arguing that Iraq had minimized its WMD holdings after 1996 would have been a very controversial position indeed.

Some of the problems that led to our misunderstanding of Iraq's WMD may be insoluble, at least by bureaucratic changes. The forms of pressure exerted on the intelligence community by the Bush Administration were perfectly legal; it would probably be impossible to regulate against them. Moreover, doing so could preclude useful and necessary questioning of intelligence analysts by Administration officials. Still, some fixes do suggest themselves.

In the future we as a nation must be willing to devote enough resources to intelligence so that we will always be able to sustain a large, aggressive program to collect all manner of information and a sophisticated analysis program on all high-priority issues. In retrospect, our over-reliance on UNSCOM inspectors lulled us into a false sense of security; this in turn contributed to our inflated estimates of Iraq's WMD progress after 1998. Even though Iraq was a difficult environment for any intelligence service to operate in, and the CIA did devote substantial assets to it at all times, it would have made some difference if the Agency could have devoted still greater resources to it, even when that seemed redundant with UNSCOM's missions.

Our failings in the WMD experience also argue for a more powerful and independent director of central intelligence. The DCI currently serves at the pleasure of the President, and although he is the nominal head of the entire intelligence community, in reality he does not have much authority over most of the intelligence agencies, whose budgets and personnel come largely from the Department of Defense. The United States could make the DCI position similar to that of the director of the FBI: the President would nominate a candidate who would then need to be confirmed by Congress, and who would serve a fixed term. And the DCI could be made the true head of intelligence, with control over the budgets and personnel of all the intelligence agencies. Many of the intelligence agencies that currently report to the Secretary of Defense, including the National Security Agency and the National Reconnaissance Office, to name just two, should instead report to the DCI. These changes would put the DCI in a stronger position to resist pressure from the executive branch (or Congress) and to protect his people from the same.

Strengthening the DCI and increasing his independence might make for smarter, bolder analysis. The less intelligence analysts have to worry that the DCI is going to take heat for unpopular if accurate judgments, the more willing they will be to make them. This is not a slur against DCI George Tenet, who I think handled the difficulties of his situation extraordinarily well. But it is a recognition that DCIs must not be put in the position that Tenet was forced into.

Another step worth considering is forbidding the CIA or anyone else in government from making any intelligence estimates public for five or ten years. As someone firmly committed to the concept of open government, who believes that the CIA has benefited from its efforts in the past decade to be more open to the public, I dislike the idea of greater secrecy. However, when intelligence estimates become public, they have a huge impact on the course of foreign-policy debates, and administrations therefore find themselves with a great incentive to make sure the Agency's estimates support the Administration's preferred policy. If such estimates were not made public, an administration would have little reason to try to influence them. The government could still produce white papers, but they should come from the State Department—the agency that is, after all, officially charged with public diplomacy.

Finally, the U.S. government must admit to the world that it was wrong about Iraq's WMD and show that it is taking far-reaching action to correct the problems that led to this error. Iraq is not going to be the last foreign-policy challenge in which we must make choices based on ambiguous evidence. When the United States confronts future challenges, the exaggerated estimates of Iraq's WMD will loom like an ugly shadow over the diplomatic discussions. Fairly or not, no foreigner trusts U.S. intelligence to get it right anymore, or trusts the Bush Administration to tell the truth. The only way that we can regain the world's trust is to demonstrate that we understand our mistakes and have changed our ways.

* Kenneth M. Pollack is the director of research at the Saban Center for Middle East Policy at the Brookings Institution. He is the author of The Threatening Storm: The Case for Invading Iraq.